Volume 8, Issue 4, Spring2014

3D Grading and Pattern Unwrapping Technique for Loose-fitting Shirt

Part 1: Resizable Design Template
Abu Sadat Muhammad Sayem
The Clothing Technology Centre, The Clothing Institute, Dhaka
&
Textile Engineering
Southeast University, Bangladesh
Richard Kennon
Textiles & Paper, School of Materials
Manchester
United Kingdom
Nick Clarke
Textiles & Paper, School of Materials
The University of Manchester
United Kingdom
ABSTRACT
This paper describes the Reverse Engineering (RE) and 3D modelling techniques used to develop
a novel resizable design platform for use in virtual design; pattern flattening and automatic
grading of mens upper-body outerwear. Advanced body scanning technology and RE technique
are applied to extract a set of sectional curves. The extracted curves are further processed in a
CAD software program and used for new surface generation by applying a 3D modelling
technique to form a resizable design platform, which is intended to be used as a 3D drawing
board for the creation of virtual shirts and for the execution of 3D grading and automatic pattern
flattening. As far as can be ascertained, such a resizable design platform for combined virtual
design, automatic pattern creation and 3D grading of mens shirt has not been previously
demonstrated.

Keywords: Virtual shirt; resizable template; pattern flattening; 3D grading

Introduction decade (McCartney et al., 2000; Kim &

An emerging technique for clothing pattern
creation is that of 3D to 2D pattern
unwrapping which means the automatic
generation of 2D patterns by unwrapping or
flattening the 3D design. Notable research
has been carried out in this field in the last
Kang, 2002; Wang, Smith & Yuen, 2002;
Wang, Wang & Yuen, 2002; Sayem, 2004;
Petrak & Rogale, 2006; Petrak, Rogale &
Mandekic-Botteri, 2006; Decaudin et al.,
2006; Kim and Petrak, 2007; Fang & Ding,
2008; Fang, Ding & Huang, 2008 and
Roedel, 2008). However this technology is

Article Designation: Refereed 1 JTATM

Volume 8, Issue 4, Spring 2014
not equally advanced and sufficiently
mature for both intimate wear and outerwear
to be implemented in the industry at this
point of time (Sayem, Kennon & Clarke,
2010 and 2012). A few clothing CAD
systems such as 3D Interactive software
from TPC (Hong Kong) and 3D Runway
from OptiTex International (Israel) provide
the capability to execute pattern
unwrapping, but in a very limited context
and only for close-fitting garments. No
efficient solution is available for outerwear,
which encompasses the major portion of
clothing consumption worldwide. For
example, it was 88% of total clothing
consumption in the 27 member states of the
European Union in 2007 and worth of 260
billion (CBI, 2008).
Attempts to design virtual outerwear for the
purposes of pattern flattening were made by
Hinds and McCartney (2000), McCartney et
al. (2000), Wang, Wang and Yuen (2002),
Petrak and Rogale (2006), Petrak, Rogale
and Mandekic-Botteri (2006), Kim and
Park (2007), Yunchu and Weiyun (2007)
and Siegmund et al. (2010). All of their
works followed laborious routes to
incorporate variable ease at different places
between the virtual body and the garment
panels for outerwear. None of their
techniques consider ease distribution over
the girth measurements as practically
followed in the fashion industry during
pattern creation for outerwear.
Article Designation: Refereed
This paper presents a technique for
developing a resizable design template that
uses traditional size charts including ease
distribution over the girth measurements for
the purposes of 3D grading and pattern
flattening of loose-fitting shirt.
Methodology
The proposed resizable design template was
intended to be structured around a set of
sectional curves, which are representative of
the size parameters. These sectional curves
are extracted from a virtual model
constructed from a set of Point-Cloud data
captured from a full-body scan of a mature
male subject using NX16 body scanning
system from [TC2] (Textile/Clothing
Technology Corporation), USA. The process
followed to develop a resizable shirt
template is summarized in Figure 1 and
explained in the following sections.
CAD Model from Point-Cloud
The raw scan data processed by the NX16
analytical software as a refined body model
in .rbd file format was converted into a
triangulated body model in .obj data
format having approximately 110,000
triangles before importing into a reverse
engineering (RE) software. Using an
appropriate RE software, the triangulated
body model was converted into a CAD
model (Figure 2) having approximately
400,000 mesh triangles and this model was
further used for extracting the sectional
curves from it.
2 JTATM
Volume 8, Issue 4, Spring 2014
 Body scanning

Generate CAD model from

Point-Cloud

Slice the CAD model to extract sectional curves from

body and arms

Generate modified B-Splines to form body and sleeve shapes

of a male upper body outerwear garment

Halve the body curves vertically and Duplicate the sectional curves
generate mirror image curves of the one of one sleeve as a mirror
halves to create symmetrical curves image to the other side

Scale the body and sleeve curves

Generate new surfaces from the scaled curves to develop the

Resizable Shirt Template

Figure 1. Workflow for Resizable Shirt Template

Article Designation: Refereed 3 JTATM

Volume 8, Issue 4, Spring 2014
 Point-Cloud Polygonal Phase CAD Model

Figure 2. Virtual Model of a Male Subject

Sectional Curve Extraction torso, as can be seen in Figure 3. The

i) Curves from torso
Seven sectional curves namely neck girth,
lower neck girth, shoulder girth, chest girt at
armpit, chest girth at the fullest area, waist
girth and hip girth were extracted from the
European Standard for size designation of
clothes EN 13402-1:2001 was consulted
to identify the landmark positions of the four
primary sectional curves representative of
the neck girth, chest girth, waist girth and
hip girth.

Figure 3. The Extracted Sectional Curves

Article Designation: Refereed 4 JTATM

Volume 8, Issue 4, Spring 2014
The measurements of waist girth and hip
girth are not directly used in the traditional
pattern cutting system for mens shirts
(Aldrich 1990). However, the position of the
waist line is important for shaping and
tailoring the middle part of the body of shirts
and jackets (Aldrich 1990). The waist girth
curve was selected in this work with the aim
of maintaining an option to produce a
tailored silhouette of mens shirts on the
proposed design template; and the hip girth
curve was selected as its lower end contour.
In addition to the four primary sectional
curves, three secondary curves (two for the
neck to shoulder area and one for the chest
area) were found to be necessary to
accurately reproduce the geometry of upper
body garments. These girth measurements
are not traditionally used as size parameters,
because they do not correspond with easily
identifiable anatomical landmarks. As a
result, no available standard describes their
positions on the torso. For rebuilding the
neck to shoulder area, two secondary girth
curves, namely the lower neck girth and the
shoulder girth were incorporated. The
shoulder girth curve was located at
approximately 1 cm below the crown of the
shoulder on either side of the torso. The
lower neck girth is designated as being 3 cm
above the shoulder girth. To rebuild the
geometry of the chest area, a second girth
measurement at the fullest chest area was
selected.
ii) Curves from arm
In order to develop the sleeves of the
proposed 3D template, four curves, namely:
the upper arm girth, the arm girth on bicep,
the forearm girth and the wrist girth were
used. Of these, only the wrist girth is
sometimes used as a size parameter in
traditional pattern cutting systems (Aldrich,
1990). The other measurements do not
correspond to easily-defined anatomical
landmarks; they do not lend themselves to
repeatable manual measurement.
Consequently, no size standard describes the
Article Designation: Refereed
position of the arm girth measurements.
However, with the availability of body scan
data, it is possible to specify and take
advantage of non-standard anatomical
measurements, as they can offer useful
secondary measurements. The forearm girth
was located on the fullest area below the
elbow, and the upper arm girth on bicep was
located at approximately 5 cm above the
elbow joint. Later, an additional curve,
termed here as curve TS, was extracted
from the shoulder curve to define the upper
edge of the each sleeve, as can be seen in the
Figure 5.
Modified Curves Generation
The sectional curves, which are stored as an
IGES data format, were imported into an
advanced 3D CAD software package.
Naturally, these curves represent the surface
geometry of a scanned subject, but do not
necessarily provide a satisfactory shape for
the surface geometry of an outerwear
garment. Furthermore, they were found to be
broken in some places due to limitations of
the scanning and modelling processes. Using
the curve drawing tool of the CAD system,
continuous B-spline curves were generated
from each of the sectional curves by
removing the irregularities in the extracted
curves such as concave areas due to the
natural geometry of the human body (for
example the spinal cord line at the back of
the torso) and by taking the geometry of
mens upper body outerwear into account.
While drawing the modified neck girth, the
protruding parts at the front due to the
laryngeal prominence were avoided as well
as the concave segments both at the front
and back which were smoothed out. The
lateral concave segments of the lower neck
girth curve were also smoothed out in the
newly drawn modified curve. These
modified curves are illustrated in Figure 4.
Similarly modified curves for the chest girth
at the fullest area, waist girth and hip girth
are drawn.
5 JTATM
Volume 8, Issue 4, Spring 2014
 Extracted neck
Newly drawn
girth curve
neck girth
curve
Figure 4. Modified Neck Girth and Lower Neck Girth Curves Generation
Symmetrical Curves Generation
The human body is not a symmetrical
object, so the curves that are derived from it
inevitably lack symmetry. Even the newly
drawn modified curves inherit visible
asymmetry from their parent curves.
However, mass-produced clothing is
expected to have a symmetrical structure if
intentional asymmetry has not deliberately
introduced by designers. This required the
curves to be modified to meet the purpose.
The body curves were split into two halves
based on a vertical plane, and one set of
halves (the left halves) are mirror-duplicated
Figure 5. Symmetrical Body and Sleeve Curves
Waist and Hip Curves Selection
Before scaling the curves to ensure the
resizeability of the proposed shirt template,
these symmetrical curves were first tested by
Article Designation: Refereed
Newly drawn
Extracted
lower neck girth
lower neck
curve
girth curve
to the right side, and each of the left halves
were joined with its respective mirrored
curves to form completely closed and
continuous curves. This generates fully
symmetrical body curves (see Figure 5).
These symmetrical curves were further used
in the subsequent steps of the process.
In order to ensure the symmetry of the
proposed 3D design template, the modified
sleeve curves from the left side are mirrored
to build the respective curves for the other
side, hence creating symmetrical curves for
the right and left sleeves (Figure 5).
Curve TS
within the
Shoulder Curve
generating a surface out of them. The
surface generation process is further
discussed in a later section. The qualitative
ability of the body curves to form
appropriate geometry of a mens shirt is
6 JTATM
Volume 8, Issue 4, Spring 2014
dealt here. A close review of the curves
representing the chest girth at the fullest
area, waist girth and hip girth; reveals
significant differences in their shape and
depth. They do not produce a good body
shape to represent as a standard shirt
geometry, as can be seen in Figure 6. The
resultant shape from these three curves has a
Figure 6. Curve Replacement and the produced Shape
For the scaling and final surface generation
processes, the duplicated copies of the chest
curve were used instead of the waist and hip
curves and they will be hereafter described
as waist and hip curves due to the locations
in which they are used.
The Scaling Process
The scaling process forms the basis of
automatic 3D grading for the parametric 3D
shirt template which can be used for the
creation of a virtual shirt and for the
extraction of 2D patterns pieces using a
flattening technique. The scaling procedure
Article Designation: Refereed
convex outline at the front and a concave
outline at the back, when seen from the
sides. To address this problem, the waist
curve and hip curve were replaced by
duplicated copies of the chest curve.
Because they utilize the same curve
geometry, they form a uniform body shape
as can be seen in Figure 6.
involves the creation of two length
parameters for each curve, selection of a
scaling point for each curve and inputting
one or more scaling factors for each curve.
As a part of the scaling process, all of the
modified and symmetrical body and sleeve
curves are first defined as length parameters.
Similarly, for each of the curves, an
additional parameter is defined to facilitate
the acceptance of external values from a size
table for the purpose of resizing or grading.
Designations of the length parameters are
summarized in Tables 1 and 2. To make the
scaling process functional it is necessary to
link the set of second length parameters with
7 JTATM
Volume 8, Issue 4, Spring 2014
external size databases for importation of necessary values whenever required.

Table 1 List of Parameters and Scaling Factors for Body Curves

Parameter
Parameter
Names of The Designation for
Designation for Scaling Factor (A'/A)
Curves External Values from
Existing Length (A)
Size Table (A')
1 Neck Girth Ng Ng_ST Ng_ST/Ng
Lower Neck
2 LNg LNg_ST LNg_ST/LNg
Girth
3 Shoulder Girth Sg Sg_ST Sg_ST/Sg
Chest Girth at
4 Cg_ap Cg_ap_ST Cg_ap_ST/Cg_ap
Armpit
Chest Girth at
5 Cg Cg_ST Cg_ST/Cg
the Fullest Area
6 Waist Girth Wg Wg_ST Wg_ST/Wg
7 Hip Girth Hg Hg_ST Hg_ST/Hg
Centre Length
8 CL CL_ST CL_ST/CL
(Shirt Length)

Table 2 Lists of Parameters and Scaling Factors for Sleeve Curves

Parameter
Parameter
Names of The Designation for
Designation for Scaling Factor (A'/A)
Curves External Values from
Existing Length (A)
Size Table (A')
1 Curve TS Ts Ts_ST Ts_ST /Ts
Upper Arm
2 UAg UAg_ST UAg_ST /UAg
Girth
Arm Girth at
3 Ag_b Ag_b_ST Ag_b_ST /Ag_b
Bicep
4 Forearm Girth Fag FAg_ST FAg_ST /Fag
5 Wrist Girth Wrg Wrg_ST Wrg_ST /Wrg
6 Sleeve Length SL SL_ST SL_ST/SL

The scaling factors for each of the curves are
defined as A'/A as can be seen in Tables 1
and 2, where:
A' = the value of a desired size of
girth curve; and
A = the existing circumferential value
of that girth curve.
The value of the desired size for each curve
equates to the size measurement for a
particular type of clothing.
The subsequent step involves selecting
scaling points for each of the curves from
which they enlarge or diminish themselves;
it was also necessary to incorporate into the
software a scaling factor, so that the
program would be able to determine the
extent of the enlargement and diminution of
each curve during resizing and grading. To
implement this process within the CAD
system in use, the scaling options include

Article Designation: Refereed 8 JTATM

Volume 8, Issue 4, Spring 2014
scaling from point and/or scaling with
three factors.
i) Scaling the Body Curves
For the purpose of scaling, the body curves
were divided into two groups: a) shirt
length-independent curves (neck, lower
neck, shoulder, chest and waist curves), and
b) length-dependent curves (hip curve). As
clothing size tables are developed within the
standard height groups (Aldrich 1990), it is
assumed that the neck, lower neck, shoulder,
chest and waist curves, which are not
dependent on the shirt length, require
scaling only in the horizontal plane.
However, for the hip curve, it is also
necessary to scale in the vertical plane to
ensure that the change in shirt length is
executed as per designers and customers
requirements.
The middle point of the lines joining the
front and back part of each curve, as can be
seen in Figure 7, were selected as scaling
points for all the body girth curves except
the hip curve. Designations of two
Scaling Factor (A'/A) for the Neck Curve,
= Ng_ST/Ng = 450/402.716540702 1.12
Article Designation: Refereed Figure 7.
Scaling
parameters for each of the body curves are
listed in Table 1. The body curves, with the
exception of the hip curve, are scaled only in
the horizontal plane (i.e. + X and + Y axis)
with a single factor for each of the curves.
For the neck curve, for example, Ng
represents its original length which was
402.72 mm; and Ng_ST represents the value
for resizing it, which would be input through
an Excel size table. If the value of Ng_ST
was 450mm, for example, the scaling factor
(A'/A) would be 450/402.72 (=1.12). Using
the appropriate scaling tool from the 3D
CAD system in use, the neck curve would
be scaled from the pre-determined point,
shown in Figure 7, with a factor of
Ng_ST/Ng, which was in this case 1.12.
However, the value of the scaling factor is
variable depending on the value of Ng_ST
from the size tables. Similarly all of the
body curves except the hip curve are scaled
with their respective factors as expressed in
Table 1.
Scaled Neck Girth Curve
Scaling Point
Original Neck Girth Curve
9 the Neck Curve
JTATM
Volume 8, Issue 4, Spring 2014
The hip curve defines the lower end contour
of the proposed shirt template. As the length
of the mens shirt can vary with the size
change or within the same size depending on
the design, the position of the scaled hip
curve should vary in the vertical plane (i.e.
according to the change in shirt length). In
order to relate the hip curve with the shirt
length, the shirt length is first defined as a
Figure 8. Scaling the Hip Curve
The lower neck curve is essentially used to
reproduce the neck to shoulder geometry
appropriately. However, this is not a
standard measurement location from the
clothing manufacturing point of view. To
reduce the number of parameters in the size
table, this girth measurement is not
considered as a direct size parameter for the
proposed shirt template. Rather it is linked
to the neck girth using the following
expression:
Article Designation: Refereed
vertical line from the center of the neck to
the hip curve, as shown in Figure 8, and
scaled with a factor (A'/A) of CL_ST/CL as
described in Table 1. The hip curve is then
scaled from the upper end point of the scaled
center length line as shown in Figure 8 and
with the factor CL_ST/CL for the +Z
direction and with the factor Hg_ST/Hg for
the + X and + Y directions.
LNg_ST = Ng_ST x LNg/Ng
So the system will calculate the value of
LNg_ST (the value of the lower neck girth
used for resizing the template) based on the
value of Ng_ST (the value of the neck girth
from the size table) as a factor of LNg/Ng
(the ratio of the original length of the lower
neck girth to the original length of the neck
girth). As a result the size change of the
lower neck girth will always be dependent
on the size change of the neck girth.
10 JTATM
Volume 8, Issue 4, Spring 2014
ii) Scaling the Sleeve Curves the front and back parts of the chest girth at
the armpit level and the shoulder girth
Similarly to the body curves, the sleeves
curves respectively. The curve TS and the
curves are also divided into two categories,
upper arm curve were scaled on the
denoted as: a) sleeve length-independent
horizontal plane based on the factors
curves (curve TS, upper arm, arm girth at
described in Table 2. Similarly, the curves
bicep, forearm curve); and b) sleeve length-
representing arm girth at the bicep and the
dependent curves (the wrist curve).
forearm girth were scaled on the horizontal
Designations of two parameters for each of
plane, but the scaling points were taken as
the sleeve curves are listed in Table 2. For
the middle points of the lines joining the
the upper arm curve and for the curve TS
curves from the left and right sleeves, as can
representing the upper edge of the sleeve,
be seen in Figure 9.
the middle points of the halving lines joining

Scaling point for sleeve line

Scaling point
. for 2nd scaling
of wrist curve
.
Sleeve line

Scaling point for arm curve at bicep

Scaled sleeve line

Scaling point for forearm curve

1st scaled Scaling point

wrist curve for 1st scaling
of wrist curve

Step 1
2nd scaled
wrist curve

Step 2
Figure 9 Scaling of the Wrist Girth Curve

To ensure that the change in sleeve length
corresponds with the change in clothing size
or design, the wrist curve was scaled along
both the horizontal and vertical planes. The
procedure was completed in two steps,
which are illustrated in Figure 9. Initially,
the wrist curve was scaled in the horizontal
plane from the middle point of the line
joining the left and right wrist curves. The
newly scaled wrist curve was scaled again in
the vertical plane, based on using the end
point of the scaled sleeve line as a scaling

Article Designation: Refereed 11 JTATM

Volume 8, Issue 4, Spring 2014
point, as can be seen at the left side of
Figure 9. The twice-scaled wrist curve (the
second scaled in Figure 9 was finally used in
the surface generation process described in
the next section. As a result, the scaled wrist
curve changes its size based on the change
of its value in the size table and also changes
its position based on the change in sleeve
length in the size table.
Similarly to the creation of the lower neck
curve described in previous section, the
curve TS does not represent a standard
measurement location from the clothing
manufacturing point of view. To reduce the
number of parameters in the size table, it is
linked to the shoulder girth using the
following expression:
Ts_ST = Ts/Sg x Sg_ST
So the system will calculate the value of
Ts_ST based on the values of the parameters
Ts, Sg and Sg_ST, where the value of
Sg_ST will be input from size tables and the
values of Ts and Sg are presented to the
system as length parameters.
Developing Size Databases for 3D Grading
A database of sizes from 37 to 45 for mens
shirts has been developed using the MS
Excel spreadsheet program, as detailed in
Table 3. The values for neck girth, chest
girths, waist girth, hip girth and wrist girth
were taken from the standard body
measurements for mature figures of standard
males described by Aldrich (1990) in the
second edition of the book Metric Pattern
Cutting for Menswear. The values for the
shoulder girth, upper arm girth, arm girth at
the bicep and forearm girth are not available
in any traditional size tables. For the
shoulder girth, upper arm girth, arm girth at
the bicep and forearm girth, the
measurements extracted from the scan data,
which have been used throughout this work,
were selected for use in the size table as the
measurement for size 42, because the
Article Designation: Refereed
subject from which the scan data were
generated and from which the subsequent
body model was developed, had a chest
measurement and a waist measurement close
to that described by Aldrich (1990) for this
size.
The size table was split into sub-databases
for individual sizes with appropriate ease
allowances. A minimum functional ease of
7.5 cm for the chest area was selected
following the findings of Moll and Wright
(1972) and a design ease of 8.5 cm was
selected to make a total ease of 16 cm which
corresponds to the suggested ease around the
chest (without a seam allowance) for easy-
fitting mens tee shirts and knitwear, by
Aldrich (1990).
When creating a straight side-seam
silhouette for a mans shirt, sufficient ease
allowances for the waist girth and hip girth
were selected to keep the final
measurements equal across the chest area.
Although no circumferential expansion takes
place in the arms as it does for the chest and
waist areas due to the breathing mechanism,
sufficient functional ease is required in the
sleeves to allow for unhampered arm
movement. The size parameters with ease
allowances for size 42 are included in Table
4. The same ease allowances have been
included for other sizes, namely 37, 38, 39,
40, 41, 43, 44 and 45 for this work.
However, it has to be noted that the size
databases and values of ease for each
measurement point may be altered in the
databases to meet the designers and
customers requirements. Once all the size
tables with ease allowances had been
prepared in Excel spreadsheets, they were
programmed, by assigning parameter
designations to the corresponding cells
containing size parameters, to be linked with
the 3D grading process using the Excel
link facility within the 3D CAD software
suite used in this research so that they can be
imported, as required, to support automatic
grading.
12 JTATM
Volume 8, Issue 4, Spring 2014
 Table 3 Size Table for Mens Tee-Shirt
Measurements for Mens Shirts (without ease) in cm
Size Parameter (neck size as size designation)
37 38 39 40 41 42 43 44 45
1 Neck Girth 37 38 39 40 41 42 43 44 45
2 Shoulder Girth 87 89 91 93 95 97 99 101 103
3 Chest girth at armpit 88 92 96 100 104 108 112 116 120
4 Chest girth at the fullest area 88 92 96 100 104 108 112 116 120
5 Waist girth 74 78 82 86 90 98 102 106 110
6 Hip girth 92 96 100 104 108 114 118 122 126
7 Upper arm girth 30 31 32 33 34 35 36 37 38
8 Arm girth on bicep 29.5 30.5 31.5 32.5 33.5 34.5 35.5 36.5 37.5
9 Forearm girth 27 28 29 30 31 32 33 34 35
10 Wrist girth 22 22.5 22.5 23 23 23.5 23.5 24 24
11 Shirt Length 76 78 80 81 81 82 82 82 82
12 Long Sleeve Length 87 88 88 89 89 90 90 90 90

Table 4 Size Parameters with Ease Allowance for Size 42

Clothing
Measurement Functional Design
Size Parameter measurement
without ease (cm) Ease (cm) Ease (cm)
(cm)
Neck Girth 42 0 0 42
Shoulder Girth 97 0 1 98
Chest girth at armpit 108 7.5 8.5 124
Chest girth at the fullest area 108 7.5 8.5 124
Waist girth 98 7.5 18.5 124
Hip girth 114 0 10 124
Upper arm girth 34 7 0 41
Arm girth on bicep 33 6.5 0 39.5
Forearm girth 30.5 4.5 0 35
Wrist girth 23.5 2 0 25.5
Shirt Length 82 - - 82
Long Sleeve Length 90 - - 90

Surface Generation to form the 3D Shirt
Template
Once all the body and sleeve curves have
been scaled, a geometric modelling
technique was applied to generate a new
surface out of them, employing the curve
matching and parametric synchronization
options available within the 3D modelling
function of the 3D CAD software in use, as
illustrated in Figure 10. The newly-
generated surfaces for the body and sleeves
form the desired shirt template as 3D design
platform for virtual clothing creation, pattern
flattening and automatic grading in 3D as
can be seen in the Figure 11.

Article Designation: Refereed 13 JTATM

Volume 8, Issue 4, Spring 2014
 Figure 10. New Surface Generation out of Scaled Body and Sleeve Curves

Figure 11. The Resizable Shirt Template

Conclusions
In this part of the research, a process of
developing a 3D shirt template by applying
reverse engineering and geometric
modelling techniques, and a novel scaling
process to make it ready for use in 3D
grading have been demonstrated. Its
application in virtual design and automatic
pattern flattening are described in part 2 of
this paper. This shirt template is intended to
be used as a 3D drawing board for fashion
designers to create a virtual shirt directly in
the 3D environment. The template is scaled
with 12 size parameters (7 for body and 5
for sleeves) represented by construction
curves. The scaling process forms the basis
of the parametric 3D shirt template which
can used to facilitate automatic 3D grading.
The value of the size parameters i.e. the
length of the scaled curves can be altered
individually or in group (representing the
arm, or the torso for instance) and this
procedure may be invoked whenever it is
deemed to be necessary, by the designer or
by a pattern technician.

Article Designation: Refereed 14 JTATM

Volume 8, Issue 4, Spring 2014
Acknowledgement
We are thankful to the Association of
Commonwealth Universities (ACU) in the
UK, Lectra (UK), MD3D Ltd (UK) and
Shape Analysis Ltd (UK) for technical
support and friendly co-operation.
References
Aldrich, W. (1990). Metric pattern cutting
for menswear, 2nd edition, Blackwell
Scientific Publications, Oxford.
CBI (2008). The Outerwear Market in the
EU. Rotterdam: Centre for the
Promotion of Imports from Developing
Countries (CBI).
Decaudin, P.; Julius, D.; Wither, J.;
Boissieux, L.; Sheffer, A. and Cani, M.
(2006). Virtual garments: A fully
geometric approach for clothing design.
Eurographics 2006, 25 (3), pp. 625-634.
Fang, J. and Ding, Y. (2008). Expert-based
customized pattern-making automation:
Part I. Basic patterns. International
Journal of Clothing Science and
Technology, 20 (1), pp. 26-40.
Kim, S.M. and Kang, T. J. (2002). Garment
pattern generation from body scan.
Computer-Aided Design, 35, pp611-618.
Kim, S. and Park, C. K. (2007). Basic
garment pattern generation using
geometric modelling method.
International Journal of Clothing
Science and Technology, 19 (1), pp. 7-
17.
McCartney, J.; Hinds, B.K.; Seow, B.L. and
Gong, D. (2000). Dedicated 3D CAD
for garment modelling, Journal of
Materials Processing Technology, 107,
pp. 31-36.
Moll, J.M.H. and Wright, V. (1972). An
objective clinical study of chest
expansion. Annals of the Rheumatic
Diseases, 31, pp. 1-8.
Petrak, S. and Rogale, D. (2006). Systematic
representation and application of a 3D
computer-aided garment construction
method, Part I: 3D garment basic cut
construction on a virtual body model.
International Journal of Clothing
Article Designation: Refereed
Science and Technology, 18 (3), pp.
179-187.
Petrak, S.; Rogale, D. and Mandekic-
Botteri, V. (2006). Systematic
representation and application of a 3D
computer-aided garment construction
method Part II: spatial transformation of
3D garment cut segments. International
Journal of Clothing Science and
Technology, 18 (3), pp. 188-199.
Roedel, H. (2008) Entwicklung
parametrischer virtueller Formkrper
weiblicher Personen zur
dreidimensionalen Konstruktion von
krpernaher Bekleidung und
Miederwaren [Development of
parametric virtual female bodies for the
three dimensional construction of close-
fitting and foundation garments]. AiF-
Forschungsproject Nr. 14540BG,
Research Report, Dresden: Technishe
Universitaet Dresden.
Sayem, A.S.M. (2004). Beitrag zur
Entwicklung parametrischer virtueller
Formkrper weiblicher Personen zur
dreidimensionalen Konstruktion von
krpernaher Bekleidung und
Miederwaren, Masterarbeit Nr. 1275.
Thesis (MSc.), Technische Universitaet
Dresden (TU Dresden), Germany.
Sayem, A.S.M., Kennon, W.R. and Clarke,
N. (2010). 3D CAD systems for the
clothing industry. International Journal
of Fashion Design, Technology and
Education, 3 (2), pp. 4553.
Sayem, A.S.M., Kennon, W.R. and Clarke,
N., (2012). Resizable trouser template
for virtual design and pattern flattening,
International Journal of Fashion
Design, Technology and Education, 5
(1), pp. 55-65, URL:
http://dx.doi.org/10.1080/17543266.201
1.614963
Siegmund, J.; Krywinsky, S.; Paul, L. and
Kleben, V. (2010). 3D design for
protective workwear and mens
outerwear. Textile Network, pp. 11-12,
Retrieved January 2014 from
http://www.textile-network.com/11-12-
2010-november-december/artikel_3d-
15 JTATM
Volume 8, Issue 4, Spring 2014
 design-for-protective-workwear-and-
men%E2%80%99s-outerwear-
_13772_9_924_en.html
Wang, C.C.L; Smith, S. S. and Yuen, M.F.
(2002). Surface flattening based on
Article Designation: Refereed
energy model. Computer-Aided Design,
34, pp. 823-833.
Wang, C.C.L.; Wang, Y. and Yuen, M.M.F.
(2002). Feature based 3D garment
design through 2D sketches. Computer-
Aided Design, 35, pp. 659672.
16 JTATM
Volume 8, Issue 4, Spring 2014